Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

Dataset: 80% C4 (Web Text) and 20% Python Code

Neuron 624 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.1139. Min Act: -0.0001

Data Index: 352211 (C4 (Web Text))

Max Activating Token Index: 391

Text #1

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.1099. Min Act: -0.0001

Data Index: 659761 (C4 (Web Text))

Max Activating Token Index: 86

Text #2

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.1027. Min Act: -0.0000

Data Index: 391710 (C4 (Web Text))

Max Activating Token Index: 280

Text #3

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0893. Min Act: -0.0001

Data Index: 803009 (C4 (Web Text))

Max Activating Token Index: 693

Text #4

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0979. Min Act: -0.0001

Data Index: 901204 (C4 (Web Text))

Max Activating Token Index: 210

Text #5

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0917. Min Act: -0.0001

Data Index: 943123 (C4 (Web Text))

Max Activating Token Index: 877

Text #6

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0858. Min Act: -0.0001

Data Index: 1283579 (C4 (Web Text))

Max Activating Token Index: 897

Text #7

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0924. Min Act: -0.0001

Data Index: 336260 (C4 (Web Text))

Max Activating Token Index: 227

Text #8

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0871. Min Act: -0.0000

Data Index: 701670 (C4 (Web Text))

Max Activating Token Index: 534

Text #9

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0890. Min Act: -0.0001

Data Index: 1317369 (C4 (Web Text))

Max Activating Token Index: 994

Text #10

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0857. Min Act: -0.0001

Data Index: 1012795 (C4 (Web Text))

Max Activating Token Index: 389

Text #11

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0854. Min Act: -0.0000

Data Index: 1097920 (C4 (Web Text))

Max Activating Token Index: 502

Text #12

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0858. Min Act: -0.0000

Data Index: 720885 (C4 (Web Text))

Max Activating Token Index: 283

Text #13

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0827. Min Act: -0.0001

Data Index: 157758 (C4 (Web Text))

Max Activating Token Index: 936

Text #14

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0817. Min Act: -0.0000

Data Index: 489384 (C4 (Web Text))

Max Activating Token Index: 740

Text #15

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0822. Min Act: -0.0001

Data Index: 28655 (C4 (Web Text))

Max Activating Token Index: 397

Text #16

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0848. Min Act: -0.0001

Data Index: 1417964 (Python Code)

Max Activating Token Index: 228

Text #17

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0841. Min Act: -0.0001

Data Index: 1130115 (C4 (Web Text))

Max Activating Token Index: 416

Text #18

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0814. Min Act: -0.0001

Data Index: 129577 (C4 (Web Text))

Max Activating Token Index: 168

Text #19

Max Range: 0.1139. Min Range: -0.1139

Max Act: 0.0794. Min Act: -0.0001

Data Index: 145465 (C4 (Web Text))

Max Activating Token Index: 736